People sometimes ask me for examples of where I’ve changed my mind. One example is lead. I used to be skeptical of the claims that lead exposure had major effects on behavior and intelligence. I was exposed to lots of lead when I was young, much more than most other kids I knew. And I didn’t notice any ill effects.
Of course that sort of anecdotal evidence is not very persuasive, and I’ve come around to the view that lead exposure probably did do a lot of damage to earlier generations.
Noah Smith has an excellent twitter thread discussing the amazing decline in violence since the early 1990s. Some of his commenters mention the internet, but that can’t be the whole story. Much of the decline in violence occurred before 2000. I’m more persuaded by Smith’s pointing to the reduction of lead exposure. While this theory seemed implausible to me when I first heard it, today it seems like the explanation that best fits the facts.
As far as property crime, I think there are other factors at work. In 1990, I used to get $500 each time I went to an ATM machine, which I spent gradually over a few weeks. So on any given day I was usually carrying roughly $300 in my wallet. Now I tend to carry about $40 and I use credit cards for most things.
I used to worry that my expensive TV might be stolen by a burglar. By 2017, that 52-inch Samsung TV was almost worthless, so much so that when I got a new TV last year I simply gave the Samsung to the deliveryman. I worry far less about theft than when I was in college, when I feared that my Pioneer stereo system might be stolen. The market for used electronics is no longer very robust. The same is true of antique furniture, and many other items. We are moving away from an economy based on things, and toward one based on experiences. The exception is real estate, but that can’t be easily stolen. (Are iPhones another exception?)
So that’s my theory. Less violent crime because of the removal of lead, and less property crime because there is less stuff worth stealing.
PS. The Straussian reading of this post is me bragging about how spectacular a human specimen I could have been if not for all of that lead I ingested. As Simler and Hanson tell us, there’s always a selfish hidden motive for even idealistic statements, such as “I was wrong.” 🙂
PPS. I first noticed the strange calm and politeness of young people in the early 2000s, when I went to a rock concert after a long hiatus. There was no longer the feeling in the air of imminent violence. Everyone was so . . . chill.
And yet, the Economist points out that this amazing reduction is violence is associated with a move toward sheltering our children from the dangerous outdoors:
When I was a kid, we were out and about all the time, playing with our friends, in and out of each other’s houses, sandwich in pocket, making our own entertainment. Our parents hardly saw us from morning to night. We didn’t have much stuff, but we came and went as we liked and had lots of adventures.” This is roughly what you will hear if you ask anyone over 30 about their childhood in a rich country. The adventures were usually of a homely kind, more Winnie the Pooh than Star Wars, but the freedom and the companionship were real.
These changes also help me to better understand the “reactionary impulse”. Even when changes are clearly for the good (less violence), my gut instinct is to miss the good old days. I try to push back against that impulse when considering social change.
READER COMMENTS
Brett
Jan 16 2019 at 12:52pm
Some of the prime targets for theft have also become harder to steal. You can “brick” your phone remotely if it gets stolen, and apparently new cars (since the early 2000s) are almost impossible to start without the key.
MarkW
Jan 16 2019 at 4:04pm
“…and apparently new cars (since the early 2000s) are almost impossible to start without the key.”
That’s true, but there are downsides. One is an increase in armed car-jacking in some places. And car theft hasn’t gone away entirely — the thieves now use tow-trucks and leave the stripped car sitting on the ground somewhere (sans wheels).
Still, I think Scott is right about the lack of stuff worth stealing in most American homes (even in expensive neighborhoods). I used to worry about people breaking into the family cottage when nobody was there, but now I can’t imagine what they’d find worth carrying off. Worst case is they’d have a party and drink all the beer.
There was a burglary ring around here a few years back who were specifically targeting the Indian-Americans due to the likelihood of finding gold in the homes. But having valuables like that at home (or in your pocket) is really an exception now.
Mark Bahner
Jan 16 2019 at 1:28pm
If you want to talk property crimes, they will really go down when the transportation system is fully automated in about 30 years. Bank and convenience store robberies in particular will be virtually impossible.
Since no one will own a car, everyone will get around by ordering a car (transportation on demand). One can’t very well order a car to pick one up after one has robbed a bank, convenience store, or house. Also, all cars on the road will have cameras that constantly look around 360 degrees, and will (probably) maintain at least a short-term record of what they see. So there will be a record of anyone running/walking anywhere across or near a street.
P.S. Kevin Drum has done really excellent work on lead and crime…see such items as his February 1, 2018 lead and crime update for Mother Jones…which shows data for blood levels and violent crimes in the U.S., Canada, New Zealand, Australia, and Britain (formerly Great Britain ;-)).
Grant Gould
Jan 16 2019 at 1:52pm
The next great revolutions in crime will be ways to steal real-time services and real estate, probably by subverting administrative processes (SIM-cloning-via-phone-call gives a good prototype of the former, the robo-signing scandal prefigures the latter).
These will be redefined to be violent crimes so the rate of violent crime will only fall modestly.
Scott Sumner
Jan 16 2019 at 2:50pm
Brett and Mark, Good points.
Grant, So it seems we’ll be moving from blue collar crime to white collar crime.
Benjamin Cole
Jan 16 2019 at 3:19pm
I am happy to get lead out of paint and gasoline.
There has been an increase in major league baseball picture ERA’s since the 1960s.
Since that golden era of pitching, there has been an increase in the use of relief pitchers.
I conclude that the use of relief pitchers is counterproductive and actually leading to games in which more runs are scored.
Alan Goldhammer
Jan 16 2019 at 3:38pm
Scott – It was not just lead in paint but for lots of years tetra-ethyl lead was used as a gasoline additive. there was considerable inhalation of lead by those exposed to lots of vehicular traffic. Also, don’t forget the proposal by Steven Levitt in the original “Freakonomics” about the role of abortion in crime reduction.
ChrisA
Jan 17 2019 at 2:19am
Alan – I believe the abortion law relaxation as a cause of lower crime idea has been proven to be incorrect when examined in more detail.
Justin
Jan 22 2019 at 2:11pm
Source?
RPLong
Jan 16 2019 at 4:35pm
I believe incentives also play a major role in the decrease of crime. In the first half of the 20th Century, a person could still plausibly settle down into a quiet working class life, even after a stint in prison. In today’s world, having a felony conviction on your record disqualifies you from virtually everything and is essentially a sentence to the proverbial gallows.
john hare
Jan 16 2019 at 6:28pm
I worked with a lot of felons in the 80s and 90s. Now we can’t hire them and get onto the properties of many customers. So we don’t hire them for construction. This is part of our labor shortage. How do you rehabilitate someone when they are not allowed to earn a normal living?
bob
Jan 16 2019 at 5:06pm
The Weeds podcast has an excellent discussion on iodized salt. Apparently iodized salt had a huge impact on cognitive skills and success in school. I seems to me that we should devote more efforts into initiatives like lead paint and iodized salt and less time on social engineering. And you are definitely right that young people got less aggressive in the late 90s. My son goes to a large high school that wasn’t allowed to have night games in the 80s because of violence. Now it’s a great place. This is without any significant change in demographics or any other factor. The kids are just nicer now.
JayT
Jan 16 2019 at 6:22pm
People are always talking about how the US has such high prison populations, and it seems that the general public feeling is that people are locked up for too long, for not good enough reasons. However, I don’t think I’ve ever seen people point out the fact that the explosion in prison population coincides quite well with the decrease in crime.
I’ve long been a believe in the lead theory, but as I’ve looked into it more, I can’t help but wonder if the biggest change isn’t just that the people most likely to commit crimes are more likely to be in prison.
Scott Sumner
Jan 16 2019 at 7:48pm
Alan, I seem to recall that the abortion claim was later discredited. I agree about gasoline.
RP and John, Good points.
Bob, Yes, that trend really surprised me when I became aware of it.
Jay, You said:
“I don’t think I’ve ever seen people point out the fact that the explosion in prison population coincides quite well with the decrease in crime.”
I’ve made that point. Recall the NYT headline everyone likes to mock—something like “Prison population keeps growing despite fall in crime”
Michael Sandifer
Jan 16 2019 at 9:54pm
Lengthened prison sentences may also be a factor.
Mark Z
Jan 17 2019 at 2:30am
After reading the wikipedia article for “crime drop” I discovered another interesting hypothesis which a treatise in Crime and Justice claims is the best out of 17 hypotheses tested (including the lead hypothesis) at explaining the drop. The ‘winning’ hypothesis according to them is the “security hypothesis.” Basically, improved security (not just law enforcement, but home security, auto security, central locking, etc.) played the biggest role in reducing property crime (completed and attempted), which the authors argue also in turn reduced violent crime
Here’s the link if anyone is interested: http://eprints.whiterose.ac.uk/97466/1/Farrell%20Tilley%20Tseloni%202014%20Why%20the%20Crime%20Drop%20%28C%26J%20vol43%29.pdf
Floccina
Jan 17 2019 at 10:51am
That is a very interesting paper, thanks for the link.
ChrisA
Jan 17 2019 at 2:34am
The lead hypothesis is just a bit too neat and the evidence is a bit too correlative, with lots of potential confounds for me. As an example I grew up in a very rural environment where cars were pretty rare (not in US) in the 1970’s. So I don’t think I was exposed to much gasoline lead as a kid. And I remember my childhood friends as being fairly violent nasty little creatures on the whole. Now I visit this area and the kids are sweet and nice, almost naive when compared to us. We were drinking by 12 years old, and getting into all sorts of messes by the time we were 14 (including sex and resulting babies) and drug taking by 16 (weed mostly). This is literally unthinkable to today’s kids. I would say that there are multiple causes of this change – I came from a family of 6 kids, that is very unusual nowadays so with fewer children more investment can be made in them, spending more time with adults gives much more mature models for kids to emulate. Drugs are far less available especially for kids – no drug seller is going to risk selling drugs to underaged children because the legal consequences are so bad if they get caught. Schools put far more focus on anti-violence, corporal punishment is now unthinkable. In other words socialization of children into society has changed in so many ways it seems just too simplistic to blame the change on lower lead.
Mark Z
Jan 17 2019 at 4:06am
I tend to agree. Usually, vaguely attributing a big trend to a ‘a confluence of factors’ seems like a cop out of having to decide which one dominated, but in the case of declining violence, it seems to be the correct conclusion. Too many changes that could – one might say should – reduce violence were occurring simultaneously to warrant confident attribution of causality primarily to one.
Mark Bahner
Jan 17 2019 at 11:19am
One thing to consider is that lead contamination was at least a regional issue (involving many countries) if not even a global issue in the 1960s and 1970s. For example, measurements of lead even in the Greenland ice sheet show much higher lead levels in the late 20th century than in previous centuries.
I’m not saying that the regional/global lead contamination was sufficiently high to cause a problem wherever you were (especially if you were in Africa or South America), but it’s at least a possibility.
ChrisA
Jan 18 2019 at 12:31am
Mark – One issue I see in recent times is that chemical analysis and detection techniques have vastly improved so that even very very tiny contaminants can be detected, at ppb or less levels. And people are reacting to these things with the same amount of fear as for when people measured stuff 50 years ago at ppm levels. Radiation is one example. There must be a limit even for the most toxic chemicals when they don’t have a measurable biological effect. I would bet that this is true for none local lead contamination – yes it can be measured, but I find it really hard to believe it has some specific but limited effect like changing your behavior in one area only.
Mark Bahner
Feb 4 2019 at 12:26am
Hi Chris,
Sorry about the delay responding. You wrote:
Well, the average U.S. blood lead levels in the late 1960s and early 1970s were unbelievably high…I don’t think that was exclusively an urban phenomenon.
Lead in gasoline versus lead in blood in U.S.
Michael Byrnes
Jan 20 2019 at 3:46pm
Economist Jennifer Doleac reviewed a few recent papers that ingeniously add some support for the lead-crime hypothesis.
The first notes that the level of violence in American cites in the 20th century was attributable to a confluence of factors that lead to lead exposure. There are 3 main ways that people get exposed to lead: via leaded gasoline, lead paint, and lead-contaminated drinking water. This paper looked at the latter cause. What they found was that cities located near lead refinieries tended to have lead pipes, but that lead pipes are not, in and of themselves, necessarily harmful, because lead from pipes will only leach into water of the water is acidic. So this study was able to use cities with lead pipes (but non-acidic water) and cities without lead pipes (but with acidic water) as control groups (with cities having both being the experimantal group). 20 year lagged murder rates were higher in cities where residents had been exposed to lead than in those that had not.
The second looked at Rhode Island data on prechool blood lead levels and school delinquency and suspension data.This study found that children who lived near busy streets had higher blood lead levels, particularly for kids born in the early 1990s (environmental lead has declined over time since the removal of lead from gasoline in the 80s). The children in the study were all demographically similar, but those who had higher lead levels had higher rates of being suspended from school.
The third study Doleac wrote about is in some ways the most interesting, because it looked into the efficacy of lead abatement services. The way service eligiblity works is to require 2 blood tests showing lead levels above a specified threshold – those with 2 such tests receive intensive lead abatement serices and nutritional counseling. (2 tests are required because individual tests are imprecise.) The study authors compared children who had 2 above-threshold tests (and thus received services) with those who had 1 test above the threshold and a second test just below the threshold (and thus did not receive lead abatement services). What they found is that kids who received the lead abatment services had less antiscoial behavior (suspensions, absences, school crimes, and violent crime arrests) than those who did not qualify for lead abatment services.
I think the case against lead is pretty open and shut at this point.
Mark Z
Jan 21 2019 at 4:56am
Do any of these papers try to assess what fraction of the drop in crime (let’s say during the 1990s) is attributable to the lead hypothesis? I’m sure it’s difficult to put an exact number on it; but Levitt purported to find that, for example, 5% was explained by changes in policing and the Brennan Center found 5% explained by changes in incarceration. I don’t know if even those numbers have held up though.
DeservingPorcupine
Jan 17 2019 at 8:42am
I was wondering how well lead levels might explain the Flynn effect, but a quick Googling seems to reveal but one study investigating that connection. Kinda surprising.
Comments are closed.